✶Explainer06:30
Decimalization Cut Commissions but Left Hidden Trading Costs
Saluzzi separates the dramatic decline in explicit commissions and quoted spreads from implicit costs such as information leakage and market impact. Electronic trading delivered major savings around 2002–2004, but institutional execution costs later leveled off because entering and exiting large positions remained difficult.
- Electronic trading drove the largest decline in transaction costs
- Decimalization replaced fractional quotes with penny increments
- Explicit commissions differ from implicit market impact
- Institutional cost improvements largely plateaued after 2007–2008
“But the real fact, and you look at charts and you can see this transaction cost for institutional investors dropped around two thousand and two,…”
“How much leakage was there? Did you move the stock a certain amount”
#decimalization#transaction costs#electronic trading
✶Explainer09:54
How Reg NMS Turned Competition Into Market Fragmentation
Reg NMS required brokers to respect the best displayed prices across competing venues, weakening the New York Stock Exchange's centralized specialist model. Saluzzi argues that the resulting proliferation of exchanges created latency differences, routing complexity, and new opportunities for high-speed strategies.
- Trade-through protection requires access to superior displayed prices
- NYSE market share fell sharply after the rule took effect
- Multiple venues introduced quote and routing fragmentation
- High-frequency strategies expanded within the new structure
“once this this fragmentation with the sec called competition, we call fragmentation it allowed noise and different types of venues to start coming through.”
“And that's where the games began.”
#reg nms#fragmentation#exchanges#market structure
✶Explainer11:47
Why High-Frequency Trading Defies a Simple Definition
High-frequency trading is not one uniform strategy. Saluzzi identifies recurring traits—speed, technology, short holding periods, and nearly neutral net inventory—while distinguishing legitimate electronic market makers from firms that ping, sniff, or exploit venue differences.
- HFT is difficult to define as one category
- Positions are commonly offset rather than held directionally
- Holding periods are usually very short
- Market making and predatory venue arbitrage should not be treated alike
“we really didn't come up with a very good definition.”
“They're in and out very very quickly, very short holding periods exactly”
#hft#market making#inventory#trading
✶Explainer13:18
How Colocation Creates a Faster Private View of the Market
Exchanges rent server space close to their matching engines, allowing well-funded traders to process direct feeds faster. By aggregating direct exchange data themselves, those traders can construct a national best bid and offer before the slower public SIP finishes consolidating the same information.
- Colocated servers sit near exchange matching engines
- Equal cable length applies within a facility, not across the whole market
- Direct feeds can outrun the consolidated public SIP
- Fractions of a millisecond are commercially valuable
“If the trick is to be as close as possible so that you can see the data coming at or process the data as fast…”
“a half of a millisecond, So I Frequency Trader was all the time in the world.”
#colocation#latency arbitrage#sip#market data
✶Explainer20:46
The Flash Crash Was a Liquidity Failure, Not Just One Large Order
Saluzzi rejects the explanation that a single large futures order caused the 2010 flash crash. He describes an already-fragile market in which automated traders passed positions among themselves, exhausted limits, canceled quotes, and withdrew, leaving market orders to search through nearly empty books.
- Limit-order books emptied rapidly
- Instant cancellation made displayed liquidity unreliable
- Automated firms engaged in hot-potato trading
- The cited futures order was not fully executed before the crash ended
- Withdrawal by liquidity providers amplified the fall
“The cause was the market was on pins and needles. The bids and offers just evaporated.”
“Everybody just walked away.”
#flash crash#liquidity#market orders#systemic risk
✶Explainer29:50
How Dark Pools Shifted From Block Trading to Tiny Trades
Dark pools began as crossing networks where institutions could match large blocks without advertising them publicly. Saluzzi says many evolved into opaque, continuously operating venues with trade sizes below 200 shares and insufficient disclosure about participant priority and access.
- Original dark pools crossed large institutional blocks
- Modern pools often execute trades as small as lit venues
- Participants may not know who receives priority
- Lack of transparency can advantage one investor class over another
“dark pools are off exchange venues that don't need to publish their quotes.”
“They have no idea who's getting teared and who gets the order first.”
#dark pools#block trading#transparency#institutional investors
✶Explainer57:33
Why the Speed Race Shrinks Margins but Concentrates HFT Power
Every new microwave, laser, or data-processing advantage forces competitors to spend merely to remain viable. That arms race compresses margins, pushes smaller firms out, and leaves larger electronic market makers with more concentrated influence rather than eliminating HFT.
- Speed infrastructure creates compulsory competitive spending
- Infrastructure costs reduce HFT profit margins
- Smaller firms are likely to disappear
- Large firms gain market share and power
- HFT is evolving rather than going away
“It's the term they call the race to zero.”
“The bigger ones will get bigger, which does become an issue of how much power they have.”
#hft#latency#competition#market concentration